对齐与安全论文趋势 2023–2026

RLHF、偏好优化、红队与安全对齐

统计范围：NeurIPS、ICLR、ICML、CVPR 四个会议，2023–2026 年，共 4,558 篇被归入该主题的论文。

主要结论

每年论文数

各会议论文数

按年份和会议统计

对齐与安全论文数，按年份和会议
年份NeurIPSICLRICMLCVPR合计
202313887101112438
2024348156246125875
20254554733001351,363
20264806235512281,882

代表论文

按年份和会议均匀挑选，方便快速看到每一届的样貌

全部对齐与安全论文

年份全部2023202420252026
会议全部NeurIPSICLRICMLCVPR

共 1,421 篇，第 24 / 29 页

  1. NeurIPS 2024Policy Aggregation
  2. NeurIPS 2024Predicting the Performance of Foundation Models via Agreement-on-the-Line
  3. NeurIPS 2024Preference Alignment with Flow Matching
  4. NeurIPS 2024Preference Learning Algorithms Do Not Learn Preference Rankings
  5. NeurIPS 2024Pretraining with Random Noise for Fast and Robust Learning without Weight Transport
  6. NeurIPS 2024PrivacyLens: Evaluating Privacy Norm Awareness of Language Models in Action
  7. NeurIPS 2024ProgressGym: Alignment with a Millennium of Moral Progress
  8. NeurIPS 2024Protected Test-Time Adaptation via Online Entropy Matching: A Betting Approach
  9. NeurIPS 2024ProTransformer: Robustify Transformers via Plug-and-Play Paradigm
  10. NeurIPS 2024Provable Editing of Deep Neural Networks using Parametric Linear Relaxation
  11. NeurIPS 2024Provably Mitigating Overoptimization in RLHF: Your SFT Loss is Implicitly an Adversarial Regularizer
  12. NeurIPS 2024Provably Safe Neural Network Controllers via Differential Dynamic Logic
  13. NeurIPS 2024PureGen: Universal Data Purification for Train-Time Poison Defense via Generative Model Dynamics
  14. NeurIPS 2024Quantifying and Optimizing Global Faithfulness in Persona-driven Role-playing
  15. NeurIPS 2024Quantifying the Bitter Lesson: How Safety Benchmarks Measure Capabilities Instead of Safety
  16. NeurIPS 2024Query-Based Adversarial Prompt Generation
  17. NeurIPS 2024RA-PbRL: Provably Efficient Risk-Aware Preference-Based Reinforcement Learning
  18. NeurIPS 2024Rainbow Teaming: Open-Ended Generation of Diverse Adversarial Prompts
  19. NeurIPS 2024RAMP: Boosting Adversarial Robustness Against Multiple $l_p$ Perturbations for Universal Robustness
  20. NeurIPS 2024Rapid Plug-in Defenders
  21. NeurIPS 2024RashomonGB: Analyzing the Rashomon Effect and Mitigating Predictive Multiplicity in Gradient Boosting
  22. NeurIPS 2024Reconstruct and Match: Out-of-Distribution Robustness via Topological Homogeneity
  23. NeurIPS 2024RedCode: Multi-dimensional Safety Benchmark for Code Agents
  24. NeurIPS 2024RedCode: Risky Code Execution and Generation Benchmark for Code Agents
  25. NeurIPS 2024Refusal in Language Models Is Mediated by a Single Direction
  26. NeurIPS 2024Regularized Conditional Diffusion Model for Multi-Task Preference Alignment
  27. NeurIPS 2024Regularizing Hidden States Enables Learning Generalizable Reward Model for LLMs
  28. NeurIPS 2024Reimagining Mutual Information for Enhanced Defense against Data Leakage in Collaborative Inference
  29. NeurIPS 2024Relational Verification Leaps Forward with RABBit
  30. NeurIPS 2024ReMoDetect: Reward Models Recognize Aligned LLM's Generations
  31. NeurIPS 2024Reproducibility study of "Robust Fair Clustering: A Novel Fairness Attack and Defense Framework"
  32. NeurIPS 2024Reproducibility Study of "Robust Fair Clustering: A Novel Fairness Attack and Defense Framework"
  33. NeurIPS 2024Rethinking LLM Memorization through the Lens of Adversarial Compression
  34. NeurIPS 2024Revisiting Adversarial Patches for Designing Camera-Agnostic Attacks against Person Detection
  35. NeurIPS 2024Revisiting Few-Shot Object Detection with Vision-Language Models
  36. NeurIPS 2024Risk-Averse Fine-tuning of Large Language Models
  37. NeurIPS 2024Robust Conformal Prediction Using Privileged Information
  38. NeurIPS 2024Robust Prompt Optimization for Defending Language Models Against Jailbreaking Attacks
  39. NeurIPS 2024Robust Reinforcement Learning from Corrupted Human Feedback
  40. NeurIPS 2024RTify: Aligning Deep Neural Networks with Human Behavioral Decisions
  41. NeurIPS 2024Rule Based Rewards for Language Model Safety
  42. NeurIPS 2024Safe LoRA: The Silver Lining of Reducing Safety Risks when Finetuning Large Language Models
  43. NeurIPS 2024SafeSora: Towards Safety Alignment of Text2Video Generation via a Human Preference Dataset
  44. NeurIPS 2024Safetywashing: Do AI Safety Benchmarks Actually Measure Safety Progress?
  45. NeurIPS 2024SafeWorld: Geo-Diverse Safety Alignment
  46. NeurIPS 2024SampDetox: Black-box Backdoor Defense via Perturbation-based Sample Detoxification
  47. NeurIPS 2024Scalable Neural Network Verification with Branch-and-bound Inferred Cutting Planes
  48. NeurIPS 2024Scaling Laws for Reward Model Overoptimization in Direct Alignment Algorithms
  49. NeurIPS 2024Scanning Trojaned Models Using Out-of-Distribution Samples
  50. NeurIPS 2024Secret Collusion among AI Agents: Multi-Agent Deception via Steganography

在 Claude、Cursor 里直接问这个趋势

接上 Paperlist MCP 后，写综述、挑方向或准备投稿时，不用离开对话框就能拿到这页的数字：每年各会议的论文数、增长倍数、涨得最快的子方向和代表论文，AI 会按你的问题整理。

可以直接复制这句话问 AI

用 paperlist 查对齐与安全在 NeurIPS 2023–2026 的论文趋势，哪一年涨得最快？

AI 会调用的工具：get_topic_stats {"topic": "alignment-and-safety"}

两分钟接入 Paperlist MCP（Claude Code、Cursor、Codex、ChatGPT）

常见问题

2023–2026 年四大会议一共有多少篇对齐与安全论文？
共 4,558 篇，按年份分别是 2023 年 438 篇、2024 年 875 篇、2025 年 1,363 篇、2026 年 1,882 篇。
NeurIPS 2026 有多少篇对齐与安全论文？
NeurIPS 2026 有 480 篇对齐与安全论文，上一届是 455 篇。
ICLR 2026 有多少篇对齐与安全论文？
ICLR 2026 有 623 篇对齐与安全论文，上一届是 473 篇。
ICML 2026 有多少篇对齐与安全论文？
ICML 2026 有 551 篇对齐与安全论文，上一届是 300 篇。
CVPR 2026 有多少篇对齐与安全论文？
CVPR 2026 有 228 篇对齐与安全论文，上一届是 135 篇。
哪个会议的对齐与安全论文最多？
NeurIPS，2023–2026 年共 1,421 篇。

「对齐与负责任 AI」下的其他主题